Spatiotemporal Deformable Part Models for Action Detection

By Yicong Tian, Rahul Sukthankar, Mubarak Shah. 2013

272 citations recorded by the source corpus.

Subfields: Biomedical Engineering, Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Biomedical Engineering

In Computer Vision and Pattern Recognition

Citations in the corpus

Delving into Egocentric Actions (2015, neutral): “Tian et al. [34] extend the deformable part model to 3D volumes and learn 3D spatio-temporal parts for action detection.”
Dense-Captioning Events in Videos (2017, background): “With the introduction of large scale activity datasets [26, 21, 15, 4], it has become possible to categorize videos into a discrete set of action categories [32, 13, 11, 52, 46].”
Histogram of Oriented Principal Components for Cross-View Action Recognition (2014, methodology): “Finally, we propose a global descriptor that is extracted from the normalized spatio-temporal distribution of STKs in 4-D. STK-Distribution and local HOPC descriptors were evaluated for cross-view human action recognition on the Northwestern-UCLA and UWA3D Multiview Activity II datasets.”
PKU-MMD: A Large Scale Benchmark for Continuous Multi-Modal Human Action Understanding (2017, methodology): “Most methods are designed for offline action detection [38, 31, 50].”
Finding action tubes (2014, methodology): “[38] extend the deformable parts model, introduced by [12] for object detection in 2D images, to video using HOG3D feature descriptors [22].”
Jointly Modeling Deep Video and Compositional Text to Bridge Vision and Language in a Unified Framework (2015, background): “…spent decades of efforts in video content understanding, including action classification (Wang et al. 2011; Sadanand and Corso 2012; Karpathy et al. 2014), detection (Tian, Sukthankar , and Shah 2013; Zhang, Zhu, and Derpanis 2013) and tagging (Yao et al. 2013; Moxley, Mei, and Manjunath 2010).”
Every Moment Counts: Dense Detailed Labeling of Actions in Complex Videos (2015, background): “More recent work extended latent SVMs to spatio-temporal action detection and localization [30, 13].”
End-to-End Learning of Action Detection from Frame Glimpses in Videos (2015, background): “It must learn to effectively utilize these observations, or frame glimpses, to reason on the temporal bounds of actions.”
Online Human Action Detection using Joint Classification-Regression Recurrent Neural Networks (2016, methodology): “Most methods are designed for action recognition [13,14,17], i.e., to recognize the action from a well-segmented sequence, or ofﬂine action detection [18,10,8,19].”
Action Classification with Locality-Constrained Linear Coding (2014, methodology): “The performance of our proposed algorithm is compared with ten state-of-the-art algorithms [18], [19], [10], [6], [23], [16], [8], [12], [25], [11].”
Learning to Track for Spatio-Temporal Action Localization (2015, neutral): “[39] extend the deformable part models, introduced in [6] for object detection in 2D images, to 3D images by using HOG3D descriptors [24] and employ a sliding window approach, in scale, space and time.”
Multi-region Two-Stream R-CNN for Action Detection (2016, unknown): “[6] extend the 2D deformable part model [13] to a 3D deformable part model and Wang et al.”
Watch-n-patch: Unsupervised understanding of actions and relations (2015, methodology): “We then show that our model can be used to detect which action people forgot, a new application which we call action patching. fully labeled actions in videos [23] [29] [31], or weakly supervised action labels [9] [7], or locations of human/their interactive objects [21] [35] [25].”
Video Action Detection with Relational Dynamic-Poselets (2014, background): “We compare our results with two other methods: GMM adaption method [4] (baseline) and spatiotemporal deformation part model (SDPM) (state-of-the-art method) [19].”
The THUMOS challenge on action recognition for videos "in the wild" (2016, unknown): “…( Bojanowski et al., 2014; Duchenne et al., 2009; Karpathy et al., 2014; Niebles et al., 2010; Raptis and Sigal, 2013; Tang et al., 2012 ) and localization ( Hoai et al., 2011; Ke et al., 20 05; 20 07; Pirsiavash and Ramanan, 2014; Tian et al., 2013; Yuan et al., 2009 ) in untrimmed videos.”
Joint action recognition and pose estimation from video (2015, methodology): “However, existing methods train models for the two tasks separately and combine the inference sequentially: taking pose estimation as input for action recognition[15, 34, 25, 8, 27, 24].”
Real-Time Online Action Detection Forests Using Spatio-Temporal Contexts (2016, background): “visual surveillance, pervasive health-care, and human-computer interaction. These new opportunities, however come with signiﬁcant challenges not encountered by the classical ofﬂine action recognition [46, 39, 10, 37, 44, 41, 47, 40, 36, 6, 13] approaches: Realistic OAD applications require real-time”
Temporal Localization of Fine-Grained Actions in Videos by Domain Transfer from Web Images (2015, methodology): “Tian et al. [34] proposed a temporal extension of the deformable part model (DPM) for action detection, they used spatio-temporal bounding boxes for training.”
Multiple/Single-View Human Action Recognition via Part-Induced Multitask Structural Learning (2015, methodology): “In our experiment, we compared the proposed method against two representative part-based methods, 2-D part model [40] and 3-D part model [34].”